News Release

Intervention reduced suicide attempts among at-risk emergency department patients

Peer-Reviewed Publication

Brown University

PROVIDENCE, R.I. [Brown University] -- In a clinical trial involving nearly 1,400 suicidal patients in the emergency departments of eight hospitals, a team led by Brown University and Butler Hospital psychologist Ivan Miller found that a multifaceted intervention lowered the relative risk of new suicide attempts by 20 percent.

In results published in JAMA Psychiatry, emergency department (ED) patients who received the intervention composed of specialized screening, safety planning guidance and periodic follow-up phone check-ins made 30 percent fewer total suicide attempts compared to people who received standard ED care.

"We were happy that we were able to find these results," said Miller, the study's lead and corresponding author and a professor of psychiatry and human behavior at the Warren Alpert Medical School of Brown University. "We would like to have had an even stronger effect, but the fact that we were able to impact attempts with this population and with a relatively limited intervention is encouraging."

A prevention intervention

While suicide prevention efforts such as hotlines are well known, published controlled trials of specific interventions have been much rarer, Miller said. This report was one of several from the Emergency Department Safety Assessment and Follow-up Evaluation (ED-SAFE) study led by Miller, Professor Edwin Boudreaux of the University of Massachusetts and Dr. Carlos Camargo of Massachusetts General Hospital and Harvard University. This study might be the largest intervention trial conducted so far in the U.S., Miller said. It focused on an especially high-risk group: emergency department patients who said they had engaged in suicidal ideation or had made an attempt within a week before their ED visit.

The trial took place in three phases to create three comparison groups. In the first phase, August 2010 to December 2011, 497 patients received each ED's usual treatment as a control group. In the second phase, September 2011 to December 2012, 377 patients received additional suicide screening. In the third phase, July 2012 to November 2013, 502 patients received the experimental intervention. Those patients received additional suicide screening from ED physicians, suicide prevention information from nurses and a personal safety plan that they could opt to fill out to be better prepared for times when they might begin to harbor suicidal thoughts again. Over the next year they also received brief, periodic phone calls from trained providers at Butler Hospital who would discuss suicide risk factors, personal values and goals, safety and future planning, treatment engagement, and problem solving.

The intervention was designed to directly involve a designated loved one whenever feasible, as well.

In all three phases, patients were briefly screened for suicidality at the ED and were also followed for a year with periodic assessment phone calls. Regardless of phase, patients who demonstrated a specific suicide risk during assessments were connected with the Boys Town suicide prevention hotline.

The number of suicide attempts and the proportion of people attempting suicide declined significantly in the intervention group compared to treatment as usual. The middle group, which received only additional screening, did not show a significant drop compared to the treatment as usual group.

Suicide attempts were not the only measure the researchers employed to understand the potential impact of the intervention. Fortunately, there were so few deaths by suicide among patients (only five total) that there could be no statistically valid conclusions drawn from that data point. But the researchers also created a broader suicide composite score which included not only attempts and deaths, but also interrupted or aborted attempts, and acts to prepare an attempt. Across the three groups, 46.3 percent of the patients reported one or more of these behaviors, but the relative risk declined significantly among people in the intervention compared to the usual care group (by 15 percent), but not among people who received screening alone.

While other interventions have also been found to reduce suicide risk, some of the most effective ones have involved providing patients with many hours of psychotherapy.

"This intervention was significantly less costly than most other interventions," he said. The research team is currently engaged in a cost-effectiveness analysis.

He noted that the intervention was associated with significant declines in suicide attempts, even though not every patient engaged in the full intervention (e.g. only 37.4 percent reported receiving a safety plan and nearly 40 percent did not complete a follow-up phone call). The intervention's apparent efficacy also persisted despite the study's ethical design, in which even people in the control phases received suicide prevention counseling that could have prevented an attempt if they presented an urgent need.

Continuing work

In further studies, the ED-SAFE team is looking at whether more intensive safety planning while patients are in the ED could help further. Miller and his colleagues are also conducting further tests of the phone follow-ups with patients from Butler Hospital and the Providence Veterans Affairs Medical Center.

###

In addition to Miller, Boudreak and Camargo, the paper's other authors are Drs. Sarah Arias and Richard Jones of Brown; Ashley Sullivan, Janice Espinola and Dr. Kohei Hasegawa of Massachusetts General Hospital; Dr. Michael Allen of the University of Colorado; Dr. Amy Goldstein of the National Institutes of Health; and Dr. Anne Manton of Cape Cod Healthcare. Drs. Lisa Uebelacker, Brandon Gaudiano and Lauren Weinstock of Brown University also contributed to the work.

The National Institute of Mental Health (grant: U01MH088278) funded the research.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.